Reviving Caslon

Part 1: The snare of authenticity

How much should a revival of a typeface look like the original? Well, just as with performing an old song—an analogy Matthew Carter has made—there is something you have to like in the original in order want to revive it. And you can't depart from the original too much, or you lose the charm of the old song that appealed to you in the first place. But if it is too much like the old versions, it might be stale and dated, irrelevant. So what do you keep and what do you change? And change in what way? That's the challenge every revivalist faces.

In the process of working on my own revival of Caslon—Williams Caslon—I came to two conclusions about revivals generally. First, the pursuit of authenticity is a snare and a trap. Don't go there. Second, particularly if it's an old typeface, it's going to be harder than you imagined, and you can lose your way in the process. So you'd better start with a very clear goal for your revival, and stick to it.

Here's the experience that led me to those conclusions.

It all started with an argument with the usual suspects at Typophile.com over the merits of Caslon—or lack of them. At one point type designer John Hudson wrote, "Sadly, Adobe Caslon is the only version that is suited to a wide range of typographic application, but it doesn't look like Caslon, so what's the point?"

At this point, a little bell went off in my head. I remembered reading Einstein's 'Autobiographical Notes' back in my student days, and of course had to ID the typeface—fortunately, on "i love typography" I don't have to explain why! The typeface was Mergenthaler Linotype Caslon Old Face, 12 point. And it always stuck in my head as having a distinct charm and readability that I never found quite matched in another typeface.

With a warning that visual impression at size in print and on screen are very different, here's a bit of the Einstein autobiography:

So when Hudson wrote about how Adobe Caslon didn't quite do it, I thought: maybe I can capture what I so liked in that Caslon.

So I started drawing, and meanwhile started researching Caslon. Along the way, I got the facsimile of the big 1766 specimen book of Caslon—originally published just before William Caslon the 1st died—edited with notes by James Mosley. I also got to talk with Mosley, and got hold of a number of other specimens.

The first thing I learned, which was a little startling, was that there is no such thing as a typeface called "Caslon." Caslon was, in fact, the person to produce a full range of roman and italic faces at all sizes. But he was working in the 18th century, and had no concept that different sizes had to match in design. —That idea only became established in the late 19th century. Furthermore, he was a kind of revivalist himself, taking as his models faces from different Dutch and English punch cutters.

So his different sizes have quite different designs. For example, here is his Pica 2 (12 point):

And here is a large size that Caslon did, the caps about 42 pt:

The serifs on these large characters are pretty heavily bracketed, and rather pointed, while the Pica 2 has very little bracketing, and blunt ends. The H is narrow here and wide in the Pica 2, and the C has two spurs, whereas the Pica 2 has one, and on and on. Essentially these are two different typefaces, by modern standards. And indeed, ATF's Caslon 540 is a beautifully conceived revival of some of the larger size Caslon designs. And Matthew Carter's Big Caslon is an elegant interpretation of the very largest, highest contrast sizes. However, my interest was in the high readability and charm of the text size. And here it gets even more complicated, to the point where the name "Caslon" becomes almost more of a Rorschach test than the name of a clearly identifiable design.

The letters on the left are blown up from the type in the 1766 Caslon specimen book. The letters on the right are from the Mergenthaler Linotype Caslon Old Face specimen book of 1928. Both are of the Pica size, and probably the Lino is based on the original Pica 2.

There are several differences of interest here. First the arches in mnh are all different. Which is "Caslon"? One way to get around this problem is to try to copy each one exactly, to be "authentic." More about this strategy shortly. A second obvious difference is that the ink spread is much heavier in the 18th century printing than the early 20th century. This is further complicated by the fact that even in the 20th century, with letterpress, rough or smooth paper makes a big difference. Here from the 1928 specimen book is an illustration of the difference. The same text with the same typeface printed on rough and smooth paper:

So which is "authentic"? Is it rough or smooth paper? Or should the original metal foundry type be seen as "authentic" and the digital form imitate that?

The late Justin Howes decided to do an "authentic" digital revival, and his strategy was to find old foundry type and print a wide range of sizes anew, then scan the printed letters, and digitize the outlines. The result is very impressive, particularly at large sizes, but it is not "authentic". This we can see in the following illustration.

Here we can see the comparison between Justin Howes' 12 point ITC Founders Caslon, in the center, with the original printed in the 18th century (left), and the early 20th century version, on rough paper (right). The point is that the digital version that tried so hard to be authentic in the ’90s is accurate to neither the original printing of Caslon—with coarse-grained ink on dampened paper—nor to the "hot metal" letterpress of Linotype. It clearly shows that it is a product of the 1990s.

However, this is not the only issue where authenticity is not achieved. If you look above, you see that Howes used the same damaged "h" for his model that Linotype used! But it is not authentic. In the original, shown above (left), the left stem leans forward, but is not bent under at the bottom, nor does it have a mangled left serif. Furthermore, there is a big problem with the italic. As James Mosley first noticed, some characters in the italic were recut sometime in the latter half of the 19th century. And in fact, at least in the 12 point size, they were completely recut, changing the slope angles. The Caslon company said this was all original and authentic in its 1924 specimen book, but they were not telling the truth. Thus the Howes' "authentic" Caslon is partly a copy of a 19th century recutting.

Do these variations from the original or the hot metal printed versions matter at all? Well, I think the answer is, "Yes." The reason for this is that weight and contrast of the 'normal' or 'regular' weight of a font at text sizes is a key factor affecting both the readability and the aesthetics of a font. The magical powers that Caslon was ascribed in the last part of the 19th century and first part of the 20th century applied to the kind of weight and contrast in the right example: letter press with relatively fine grained ink, on rough paper.

The ink spread or "gain" is actually a key factor here, because it operated in letter press as a kind of automatic "optical correction" for smaller sizes. As you can see above, what happens in letterpress is that ink spread added what in digital terms is a "stroke" around the letter. It is in effect a uniform addition to the thickness of all the stokes. But the stokes are not uniform in thickness. Adding a few thousandths of an inch to a fine stoke may double or triple its thickness, and while making only a slight percentage difference in a thick stem.

The graphic is from an illustration for Christian Schwartz's Houston, a newspaper face. So that ink gain is for a modern high speed newspaper web press. The old letter press ink gain is more, and more irregular, as you go back in history.

The story goes on—though I won't—as far as the irregularities that would have been involved in old printing. But the most important point is that the irregularity that was involved in Caslon's work today looks distressed, antique, or antiquarian. If that is what you want, then ITC Founders Caslon is fabulous at larger sizes, and a bit too light but still fascinating to the eye at smaller ones. But the reality is that just doesn't look good today as a font for regular use in books and magazines.

To go back to the analogy of the performance of an old song or old music score, the situation reminds me of what Ira Gershwin said when he heard Ella Fitzgerald's recording of the Gershwin song book. "I never knew we were so good until I heard Ella sing our songs." There is an "authentic" version, by the way, with George and Ira's sister singing with George at the piano. I'd be interested in hearing it, but for listening pleasure, I'm sure it doesn't hold a candle to Ella Fitzgerald.

Having labored in the Caslon vineyard for some years, I have great sympathy for Justin Howes (who died tragically young) and admiration for his monumental effort. But I think his effort proves that the goal of authenticity is misguided. What is most desirable today is not "authentic" but "classic." And in order for a classic to look classic today it needs to be changed.

The need for changes is particularly true for Caslon, which has over the centuries been controversial, with many publishers, authors and readers adoring it, and type designers and connoisseurs slamming it.

The reality is that Caslon was himself a revivalist, and in his rush to create his huge output—a cornucopia of glyphs for non-latin scripts and a full range of latin sizes—and get it out to the world there was a lot of sloppiness by modern standards. In fact, he was sloppy even by comparison with some continental type founders working at the same time. And yet, there is a magic in some of Caslon's sizes, and interesting work in all of them. There is no doubt that he had a rare gift.

Here is my late Uncle Ben Lieberman, in his Types of Typefaces (1967) summing up views of Caslon:

"[Caslon] is perhaps the most controversial face in history. Some persons consider it the greatest type ever (they have popularized a motto, 'When in doubt, use Caslon') and others think it overrated, a collection of mistakes, elusively out of keeping with everything. But—it works, is highly readable, alive, with warmth and open dignity that has no pretense whatsoever. Caslon is the prime example of a typeface in which the individual letters are nothing, but the total effect is strong and honest—the reverse of an all-star performance in which each letter has such perfection that it competes to be noticed."

Lieberman was writing on the cusp of the change to photo type, which in twenty years was to be replaced yet again by digital type. And the magical "total effect" that he wrote about didn't, in my view, survive the transition. My effort to recover that total effect in text is what I'll write about in part II.

William Berkson is a writer and type designer. Information about his new book, to be published in October, is here, and his new revival of Caslon can be found here.

[read part two]